MAVManipulation Analysis and Visualization
The Manipulation Analysis and Visualization (MAV) tool was designed to enable users to analyze extremely large and diverse multi-dimensional datasets and gain insight into the data and the phenomena that produced the data. It is intended to enable the discovery of previously unknown relationships, trends, patterns, continuities, and discontinuities in data. The research that informed the design of the MAV tool was completed in a partnership effort between Vigilant Cyber Systems and Utah State University.
The function of MAV is centered around the query intermediate representation (IR), which represents a complete dataflow, extending its capability beyond the strict definition of a query to include data types, analytics, and the source and destination of data. Logically the query IR represents a query in terms of data sources, filters, transformations, and sinks or destination views. Additionally, the IR can represent complex, nested structures and supports querying on the results of previous queries.
From a user perspective MAV contains a View Manager (VM), Views, and View Controllers that allow the user to interact with query results and generate new queries, display query results in desired visualizations, and provide a context of the analyst’s exploration. The general idea is query initiation and context (tree), query builder and elaboration, then visualization, and view interaction/ summary. MAV currently supports seven different views, but was designed to enable easy integration with additional views in the future.